Many-objective optimization problems (MaOPs), i.e., the number of objectives is greater than three, have drawn steady attention in the evolutionary multiobjective (EMO) community during the past few years and a number of efforts have been made to deal with it. The Pareto-based EMO algorithms may fail to convergence to the PF due that most of solutions in the population are non-dominated. Thus, we proposes a new dominance relation based on simplex, called simplex-dominance, in this paper. Therein, the proportion of the comparability between any two points in the objective space is enlarged, and this produces finer selection pressure towards Pareto front. The proposed algorithm is evaluated on three DTLZ problems with 10 objectives and compared with the current NSGA-III. The simulation results show that the proposed algorithm performs better on convergence than that of NSGA-III.